Product Design' Accessibility Mandate in the AI Age

We talk a lot about AI reshaping the future.

We talk less about who gets to participate in it.

In this episode of FUTUREPROOF., I sit down with Corbb O’Connor, who leads accessibility advocacy at Level Access. Corbb is blind. He’s spent years consulting enterprise teams — from financial institutions to global brands — helping them design digital experiences that are actually usable by people with disabilities.

This isn’t a compliance conversation.

It’s a systems conversation.

As AI systems increasingly generate interfaces, content, decisions, and workflows at scale, accessibility can no longer be an afterthought. If accessibility isn’t embedded upstream — in product design, in data pipelines, in AI outputs — exclusion compounds just as quickly as innovation.

Corbb argues that inclusion is not a moral add-on. It’s infrastructure. It’s economics. It’s risk management. And increasingly, it’s competitive advantage.

We explore:

•  Why accessibility should be treated like cybersecurity — a non-negotiable requirement, not a retroactive fix 
•  The difference between “AI for accessibility” and “accessible AI” 
•  Why automated scanning tools can’t replace human testing 
•  How poor product design quietly excludes users without teams even realizing it 
•  Why psychological safety and culture matter just as much as tooling 
•  And whether AI will widen or narrow accessibility gaps over the next five years 

If digital products define access to banking, healthcare, employment, and civic life, then accessibility isn’t a feature.

It’s participation.

And as AI becomes core infrastructure, the question becomes sharper:

Are we scaling inclusion — or scaling exclusion?